The recent advent of green fluorescent protein (GFP)-labeling allows one to follow the behavior of selected molecules in the living cell. With this technology the gene for a known protein is spliced with a GFP gene, and the product is then expressed in a living cell as a fusion protein. Under favorable conditions the fusion protein functions normally and its position can be followed over time in the living cell by low light level fluorescence imaging. However, long term GFP imaging studies are tricky because, like most fluorescent probes it photobleaches, and also because over time the levels of blue light needed to excite GFP can be toxic to the cell. Thus, for long term studies a low light level camera must be coupled to a high quality fluorescent imaging platform on which the light used to illuminate the specimen can be minimized and shuttered. It is also important that the system allow the specimen to also be optically sectioned in the epi-fluorescent mode, and als o followed for long periods in a non-fluorescent imaging mode. Since the BMIRR had many of the LM components necessary for GFP imaging, we spent part of 1997 assembling a GFP imaging station for internal and external use. For this we modified the DIC (De Senarmont) and fluorescence (Nikon Quadfluor) portions of our quantitative wide-field fluorescent microscope workstation. This workstation has a large format cooled CCD (PXL 14000 Photometrics Ltd.) for image acquisition, electronic shutters for time-lapse imaging, and a stepper motor for collecting Z-series (Ludl Inc.). The camera and other electronic devices are controlled by a SGI workstation running ISEE software (Inovision Corp). This microscope and workstation was purchased in 1996 entirely through NY State funding We upgraded the microscope by equipping it with a Quadfluor fluorescence system and Omega fluorescence filter cubes. The latter were designed as a collaboration between the Resource and Omega, and minimize internal reflections while providing an enhanced signal to noise ratio. With these modification's we can now obtain high resolution and high contrast images of extremely dimly labeled structures with very short (200 ms) exposures--which increases the length of time we can conduct time-lapse studies on GFP labeled cells. We also configured the platform with a pneumatic cylinder and simple electronic device (Darlinton) so that the DIC analyzer could be alternatively inserted and removed from the fluorescence light path using the ISEE software. This not only enables a maximum fluorescence signal to reach the camera, but it allows to also be imaged by DIC imaging.